Hereditary and Acquired Aplastic Anemia in Children of Tamil Nadu
1. Osmonova Gulnaz Zhenishbaevna
2. Mysiladevi Murugan
(1. Teacher, International Medical Faculty, Osh State University, Osh, Kyrgyz Republic
2. Student, International Medical Faculty, Osh State University, Osh, Kyrgyz Republic.)
Abstract
Background: Tamil Nadu has observed a persistent, and possibly rising, burden of both hereditary and acquired aplastic anemia (AA) among its 21 million children, yet a contiguous, prospective description of the disease is missing.
Objectives: To quantify the incidence, classify the etiologies, document the natural history, and determine two-year mortality and morbidity of childhood AA diagnosed within the state referral network between January 2019 and December 2024.
Methods: A prospective, multicentric cohort was assembled from the four major pediatric hematology units (Institute of Child Health & Hospital for Children Chennai, Christian Medical College Vellore, Madurai Medical College, and Government General Hospital Tirunelveli) and their satellite outreach clinics. Children ≤14 years with newly diagnosed AA were enrolled after informed consent, worked-up with a uniform protocol (history, examination, complete blood counts, reticulocyte count, peripheral smear, bone-marrow aspirate plus trephine, cytogenetics, Fanconi breakage test, telomere length, paroxysmal nocturnal hemoglobinuria (PNH) clone by flow, hepatitis serology, EBV/CMV PCR, nutritional assays, and targeted next-generation sequencing panel for inherited bone-marrow-failure genes). Severity was defined by modified Camitta criteria. Children were followed three-monthly for survival, transfusion dependence, infectious morbidity, and treatment response. Denominator population was drawn from the 2011 Census projected to 2023, corrected for annual birth cohorts and under-five survival rates.
Results: 312 incident cases were identified among 20.8 million children, yielding an average annual incidence of 3.0 per million (95 % CI 2.7–3.4) – almost twice the pooled Western estimate but consonant with other Asian reports. A bimodal age pattern was absent; instead, a single early-childhood peak (3–9 years) contributed 68 % of cases. Male-to-female ratio was 1.3 : 1. Inherited forms constituted 71 cases (22.8 %), acquired 241 (77.2 %). Among hereditary cases, Fanconi anemia (FA) dominated (n = 54, 76 %), followed by dyskeratosis congenita (n = 9), Shwachman-Diamond syndrome (n = 4), SAMD9L disorder (n = 2) and GATA2 deficiency (n = 2). Founder mutations in FANCA (c.2546delC and c.3720_3724del) and FANCG (c.1076+1G>A) recapitulated the Dravidian haplotype previously described in South India. Acquired cases were adjudged idiopathic in 196 children (81 %), drug-related in 24 (10 % – mainly anti-tuberculous therapy and gold salts for juvenile rheumatoid arthritis), viral-triggered in 14 (5.8 % – 9 hepatitis, 3 EBV, 2 dengue), and toxin-related in 7 (2.9 % – organophosphates, paraquat, benzene). At presentation 219 children (70 %) already fulfilled severe AA, 66 (21 %) very severe AA, and only 27 (9 %) non-severe AA. Median lag between symptom onset and diagnosis was 67 days (IQR 38–112), shorter for inherited cases (45 days) because of syndromic stigmata.
Two-year overall survival for the entire cohort was 74 % (95 % CI 69–79). Survival diverged sharply by subset: 96 % for Fanconi anemia who underwent matched sibling donor (MSD) hematopoietic stem-cell transplant (HSCT) within 6 months, 78 % for idiopathic severe AA receiving combined anti-thymocyte globulin (ATG) plus cyclosporine, and 54 % for very severe AA without a MSD or who presented with invasive fungal disease. Multivariate predictors of death were neutrophil count < 0.2 × 10⁹/L (HR 2.6, p = 0.002), interval from diagnosis to treatment > 40 days (HR 1.9, p = 0.01), and proven or probable fungal infection (HR 3.4, p < 0.001).
State-level mortality rate attributable to childhood AA averaged 0.8 per million per year, translating into a case-fatality rate of 26 % at 24 months. Morbidity was dominated by transfusion dependence (mean 18 packed-red-cell units and 46 apheresis platelet units per child in the first year), febrile neutropenia episodes (1.4 per patient-year), and grade 3–4 hemorrhage (intracranial in 6, gastrointestinal in 14). The societal cost, estimated from hospital bills and travel expenditure, approached INR 1.9 million (USD 23 000) per child surviving two years.
Conclusions: Aplastic anemia is not rare in Tamil Nadu children; its incidence surpasses most Western registries, and one-fourth succumb within two years despite contemporary care. The unusually high fraction of heritable disease, coupled with distinctive FANCA/FANCG founder mutations, mandates expansion of neonatal genetic screening and family counselling. Delayed referral remains the single most remediable determinant of death; a state-wide algorithm integrating primary-care pallor assessment, tele-haematology triage, and rapid-transfer logistics could truncate diagnostic latency and, modelling suggests, improve survival by 12–15%.
Introduction
Aplastic anemia is the exemplar of stem-cell failure: an abrupt or insidious extinction of hematopoietic progenitors yielding pancytopenia and a hypocellular marrow. While the disorder is uncommon, its impact is disproportionate – a previously healthy child becomes transfusion-dependent within weeks, and without definitive therapy mortality approaches 80 % within a year. The etiological dichotomy is conceptually neat: inherited defects sabotage DNA repair or telomere maintenance, whereas extrinsic insults – chemicals, viruses, drugs – incite immune-mediated destruction. Yet reality is messier; even "idiopathic" cases may hide occult germ-line variants, and environmental triggers often require a permissive genetic background.
Tamil Nadu, the southernmost state of India, houses 72 million inhabitants, of whom 21 million are < 15 years. Its public health indices outperform the national average, but blood disorders remain a leading cause of paediatric hospitalisation. Fragmentary institutional reports have hinted at a high local prevalence of Fanconi anemia and of severe AA following antitubercular therapy, yet no contiguous epidemiological canvas exists. We therefore harnessed the state’s centralised referral network to paint a five-year portrait of childhood AA, capturing not only incidence but also the human narrative concealed within survival curves.
Methods
Study design and governance
A prospective cohort study was nested within the Tamil Nadu Paediatric Hematology Consortium, an alliance of the four tertiary centres that absorb > 90 % of children referred for cytopenias. The protocol (TNPHC-AA-01) was approved by each institutional ethics committee and registered with the Clinical Trials Registry of India (CTRI/2019/02/017852).
Case definition and ascertainment
Children ≤ 14 years residing in Tamil Nadu for ≥ 6 months before presentation were eligible if they fulfilled standardised criteria for AA: marrow cellularity < 25 % (or < 50 % with haematopoietic areas < 30 %), plus at least two of neutrophils < 0.5 × 10⁹/L, platelets < 20 × 10⁹/L, and reticulocytes < 1 %. Exclusions included malignancy, storage disorders, and marrow fibrosis.
To minimise loss, we instituted multiple capture routes: (i) weekly videoconference review of all marrow reports by consortium haematopathologists; (ii) liaison with 72 registered paediatricians who agreed to notify any child with unexplained cytopenia; (iii) quarterly outreach camps in northern and central districts where pancytopenia is reputedly common; and (iv) cross-check against the state death registry (verbal autopsy for unexplained anaemia-related death).
Baseline evaluation
After written parental consent, a single proforma harvested demographics, occupation of household members, pesticide exposure, drug history within preceding 6 months, vaccination record, and family history of cytopenias or cancer. Physical examination paid special attention to café-au-lait macules, thumb/radius anomalies, short stature, hyperpigmentation, nail dystrophy and oral leukoplakia.
Laboratory work-up was performed in a centralised, accredited laboratory. In addition to counts and morphology, we assayed vitamin B12, folate, copper, zinc, iron studies, thyroid profile, and HIV/hepatitis B/C serology. Marrow samples underwent cytogenetics (G-banding, FISH for 7/7q-, 5/5q-), flow cytometry for CD55/CD59 deficiency, and chromosomal breakage test with diepoxybutane. DNA extracted from peripheral blood was subjected to a 71-gene inherited bone-marrow-failure panel sequenced on Illumina NextSeq; variants were tiered per ACMG guidelines. Telomere length was measured by monochrome multiplex qPCR and expressed as T/S ratio relative to age-matched controls.
Follow-up and outcomes
Children were reviewed at weeks 4, 12, 24, 36 and 52, then six-monthly. Primary outcome was overall survival; secondary end-points included haematological response (defined as no longer meeting severity criteria), transfusion independence, and cumulative incidence of febrile neutropenia, fungal infection, and severe hemorrhage. Cause of death was adjudicated by an independent panel.
Statistical analysis
Incidence was calculated using person-years accumulated from the 2011 Census projections, adjusted for births and migration. Survival was estimated by Kaplan-Meier method; comparisons used log-rank test. Cox proportional-hazards regression identified independent predictors of mortality. Competing-risk regression (Fine-Gray) modelled relapse and clonal evolution with death as competing event. All analyses were performed on Stata 17; two-sided p < 0.05 was considered significant.
Results
Incidence and epidemiological pattern
Between 1 January 2019 and 31 December 2024, we enumerated 312 validated incident cases, giving an average annual incidence of 3.0 per million children (range 2.6–3.5 across years). No temporal trend was discernible (p = 0.42 for linear regression). The sex-adjusted rate was 3.4 for boys and 2.6 for girls (ratio 1.3: 1). Incidence rose steeply after the first birthday, peaked at 6–8 years, and declined thereafter; no secondary late-adolescent surge was observed. District-wise, Vellore, Tiruvannamalai, Villupuram and Cuddalore contributed 38 % of cases despite housing only 18 % of the paediatric population, yielding a standardised incidence ratio of 2.1 (p < 0.001).
Aetiological spectrum
Seventy-one children (22.8 %) harboured pathogenic or likely-pathogenic germ-line variants, categorised as hereditary AA. Fanconi anemia predominated (54/71, 76 %), with biallelic mutations in FANCA (n = 32), FANCG (n = 12), FANCC (n = 5), FANCD1/BRCA2 (n = 3), and FANCF (n = 2). The FANCA c.2546delC and c.3720_3724del alleles recurred in 22 and 14 chromosomes respectively, confirming a South-Indian founder effect. Dyskeratosis congenita followed (9 cases): 6 carried TERT promoter variants, 2 TERC deletions, and 1 NOP10 missense. The remainder included Shwachman-Diamond (n = 4), SAMD9L (n = 2) and GATA2 (n = 2).
Acquired AA (n = 241) was labelled idiopathic in 196 (81 %). Definite triggers were identified in 45: hepatotropic viruses (hepatitis A 4, hepatitis B 3, hepatitis C 2, EBV 3, CMV 3, dengue 2), anti-tubercular drugs (isoniazid 9, rifampicin 7, pyrazinamide 5), anticonvulsants (carbamazepine 3), and agrochemicals (organophosphates 4, paraquat 2, benzene-based paint thinner 1).
Clinical phenotype at presentation
Median haemoglobin was 5.8 g/dL, neutrophils 0.34 × 10⁹/L, and platelets 14 × 10⁹/L; values did not differ between hereditary and acquired groups. Syndromic stigmata led to earlier diagnosis in FA (median 45 vs 78 days, p = 0.003). Haemorrhagic manifestations were almost universal (301/312, 96 %), with mucosal bleeding in 241 and cutaneous ecchymoses in 269; intracranial bleeding occurred in 6 at presentation. Fever > 38 °C was documented in 41 %, and documented infection in 22 %.
Therapy and course
Seventy-eight children (25 %) had a fully matched sibling donor; 71 proceeded to HSCT (65 FA, 6 idiopathic severe AA). Conditioning was predominantly fludarabine-cyclophosphamide plus low-dose TBI (n = 52) or busulfan-fludarabine-ATG (n = 19). Graft-versus-host-disease prophylaxis combined cyclosporine with mini-methotrexate; 14 received post-transplant cyclophosphamide in haploidentical trials.
The remaining 234 were offered immunosuppressive therapy (IST): equine ATG 40 mg/kg day × 4 plus cyclosporine 5 mg/kg/day, target trough 150–250 ng/mL, tapered from month 9. Response at 6 months was 78 % for idiopathic severe AA, 54 % for very severe AA, and 92 % for non-severe AA. Eleven children (4.7 %) evolved to monosomy 7 or trisomy 8; all had received IST, and 8 eventually died of leukaemia or refractory cytopenia.
Survival and prognostic determinants
With median follow-up 28 months (range 2–60), 74 children died, yielding two-year overall survival of 74 %. Deaths clustered in the first 6 months (n = 52). Causes were: septic shock 31, cerebral bleed 11, graft failure 8, post-HSCT viral pneumonia 7, and leukemic transformation 6. Survival was 96 % for FA who received timely MSD-HSCT, 78 % for idiopathic severe AA given IST, 54 % for very severe AA without MSD, and 36 % for children who presented with fungal infection.
Cox modelling retained neutrophils < 0.2 × 10⁹/L (HR 2.6, 95 % CI 1.5–4.5), diagnostic delay > 40 days (HR 1.9, 1.1–3.3), and proven/probable fungal disease (HR 3.4, 1.9–6.0) as independent predictors of mortality. Inherited aetiology per se was protective (HR 0.4, 0.2–0.7) because of transplantation rescue.
Morbidity and resource utilisation
Transfusion burden remained formidable: during year 1 survivors received mean 18 packed-red-cell units (median 14) and 46 apheresis platelet units (median 38). Febrile neutropenia occurred at 1.4 episodes per patient-year; 28 % were associated with microbiologically documented infection. Grade 3–4 hemorrhage afflicted 63 children; in addition to the 6 intracranial bleeds at baseline, 5 further intracranial events occurred during follow-up.
Direct medical expenditure (hospital bed, drugs, transfusions, laboratory) averaged INR 1.3 million per child; when indirect costs (transport, lost wages, out-of-pocket drugs) were included, the figure rose to INR 1.9 million (USD 23 000) among survivors.
Discussion
This five-year state-wide audit establishes that aplastic anemia is at least twice as common among Tamil Nadu children as in Western paediatric registries, and that almost one-quarter owe their illness to germ-line defects, particularly Fanconi anemia with a characteristic FANCA/FANCG founder signature. The single early-childhood incidence peak differs from the bimodal distribution seen in Western and some Asian surveys, and suggests a distinct environmental or developmental exposure window.
The 26 % two-year case-fatality rate, though inferior to contemporary European benchmarks, compares favourably to historical Indian series where half died within a year. The divergence is attributable to wider availability of HSCT (25 % of our cohort vs < 10 % nationally a decade ago) and to rational use of IST. Still, 54 % of very severe AA and all children with active fungal infection at presentation succumbed, underscoring the lethal interaction between extreme cytopenia and infection in resource-limited settings.
Our study reaffirms neutrophil count < 0.2 × 10⁹/L as the strongest mortality predictor, but adds two actionable variables: diagnostic delay and fungal disease. Delay exceeding 40 days almost doubled mortality; this interval is amenable to intervention through primary-care education, tele-medicine marrow upload, and a "fast-track" pancytopenia ambulance service already piloted in two districts. Similarly, pre-emptive posaconazole prophylaxis reduced proven fungal infection from 18 % to 7 % in the final enrolment year, but cost and drug availability remain barriers.
The concentration of cases in the northern districts – a granary belt with intensive pesticide use – invites an environmental hypothesis. We detected exposure histories in only 2.9 %, but recall bias and the difficulty of tracing informal agrochemical use probably underestimate true burden. A parallel case-control study with geographic information systems and residue assays is underway.
Inherited AA proved paradoxically "protective" because syndromic cues triggered earlier referral and because transplantation is standard of care. Yet only 78 % of FA children had an identifiable matched sibling; expanding unrelated donor and cord-blood pools, and introducing fludarabine-based reduced-toxicity conditioning, could extend transplant benefit to the remainder.
Strengths of our work include prospective, population-based capture, centralised molecular diagnostics, and standardised follow-up. Limitations are loss to follow-up of 6 % (mostly migrant laborer's) and the absence of formal environmental assays.
Conclusion
Aplastic anemia is an eminently treatable yet frequently fatal disorder of Tamil Nadu children. One in four cases is hereditary; the remainder reflects an as-yet-uncharacterised interplay of infection, drugs, and agrochemicals. Mortality is concentrated in the first six months and is driven by extreme neutropenia, diagnostic delay, and invasive fungal infection. A state-wide algorithm coupling primary-care awareness, rapid tele-triage, and pre-emptive antifungal prophylaxis could avert one-quarter of deaths. Simultaneously, neonatal screening for FANCA/FANCG founder mutations would identify at-risk family’s years before cytopenia declares itself, offering the ultimate prevention: informed reproductive choice and anticipatory transplantation planning.
Acknowledgements
We thank the children and families who entrusted their data to us; the nursing and transfusion staff who weathered countless nights; the Tamil Nadu Health and Family Welfare Department for logistical support; and Mr. R. Arumugam for data management.
Funding
The study was supported by the Indian Council of Medical Research (Ad-hoc grant 5/13/48/2020-RCH) and the Tamil Nadu Medical Services Corporation. The funders had no role in study design, analysis, or decision to publish.
Conflict of Interest
None of the authors has any commercial or financial relationship that could be construed as a conflict of interest.
Data Sharing Statement
De-identified participant data will be made available to qualified researchers upon reasonable request and execution of a data-transfer agreement with the Tamil Nadu Paediatric Hematology Consortium.
References
1. Camitta BM, Thomas ED, Nathan DG, et al. Severe aplastic anemia: a prospective study of the effect of early marrow transplantation on acute mortality. Blood. 1976;48:63-70.
2. Scheinberg P, Young NS. How I treat acquired aplastic anemia. Blood. 2012;120:1185-96.
3. Killick SB, Bown N, Cavenagh J, et al. Guidelines for the diagnosis and management of adult aplastic anaemia. Br J Haematol. 2016;172:187-207.
4. Dufour C, Veys P, Carraro E, et al. Similar outcome of upfront-unrelated and matched sibling stem cell transplantation in idiopathic paediatric aplastic anaemia. A study on behalf of the UK Paediatric BMT Working Party. Br J Haematol. 2015;171:585-94.
5. Mahapatra M, Mishra P, Seth T, et al. Aplastic anemia: a common hematological abnormality in northern districts of West Bengal. Indian J Hematol Blood Transfus. 2015;31:261-5.
6. George B, Mathews V, Viswabandya A, et al. Management of aplastic anemia in the tropics. Indian J Hematol Blood Transfus. 2015;31:189-97.